Development of a natural language processing solution for antipsychotic medication reviews.

People taking antipsychotics need their antipsychotic medication reviewed regularly to ensure their treatment is optimised. However, this is a very time-consuming process for clinicians as most of the relevant information about medication response (e.g., side effects) is recorded in unstructured free text. This project aims to develop a solution which uses natural language processing to summarise treatment history from patient records in detail to support clinicians with antipsychotic medication reviews. We will make use of the latest natural language processing tools, large language models (LLMs), as these have the power and sophistication needed for handling complex text, and have shown excellent performance on other summarisation tasks.